A great deal of research work has been carried out for many years on immobilisation of the toxic fraction of various combustible waste in a vitreous, glass ceramic or crystalline type matrix, the toxicity being either chemical (including heavy metals) or radioactive. There are often several advantages such as stabilising waste, production of a tested and durable confinement matrix, reduction of the initial volume of waste, etc.
Obviously, research for an industrial process capable of both combustion and vitrification in a single, simple unit generating only small quantities of secondary waste is economically advantageous.
Several methods have been explored in research laboratories and at industrial scales. Essentially, the use of the thermal plasmas, variants using immersed or non-immersed electrodes and melting by direct induction have been considered.
Several processes have been developed for plasma treatment, but they have disadvantages that make industrial application difficult. Crucibles are made from refractory materials and wear quickly when in contact with molten glass (by corrosion in a complex aggressive medium) and also under intense plasma radiation. The treatment capacity of combustible waste in these plasma furnaces is also limited so as to protect linings made of refractory materials.
The most frequently used plasmagenic gases are nitrogen and air. In the first case, the generated plasma is only used as a heat source and not as a combustion element, which leads to simple cracking of organic molecules. This makes the chemical composition and treatment of fumes that contain a great deal of unburned materials, soot and dust and often nitrogen oxides, very complex. The use of air as the plasmagenic gas partially solves the disadvantages mentioned above, but 80% of the gases are then useless but are still heated, which means that gas treatment units are oversized.
Plasma fusion production using cooled crucibles has been tested to overcome the problem of refractories. Copper has been proposed as a material for the construction of crucibles, but it has the same disadvantage that it is sensitive to corrosion, particularly in a nitric medium; noble stainless steel is preferred to copper. However, heat transfers from the molten material in the crucible to the walls are such that melting is always made difficult, which makes it more difficult to create a sufficiently extensive bath and to empty the crucible.
The high frequency direct induction technique for melting in a metallic crucible, at least part of which is transparent to electromagnetic fields, is also known. With this technique, melting and production of a sufficiently large glass bath and casting are controlled. Known applications include the production of high purity glass and enamels, and vitrification of high activity radioactive waste. Relevant descriptions are found in French patent applications FR 91 02595(U.S. Pat. No. 5,367,532)and FR 96 09382(U.S. Pat. No. 6,185,243). But there are disadvantages with this process if it is applied to melting of a confinement matrix above which combustible elements are thrown. In particular, there is a strong chemical interaction between the waste to be treated and the molten material, resulting in important modifications of its composition and homogeneity. For example, reduction as far as the metallic phase of an oxide-based material (glass) is practically inevitable when the waste contains carbon or hydrogen or sulphur, even if air or oxygen blowing means are used in or bath. This result modifies the required properties for the ash confinement matrix and for correct electromagnetic operation of the process. Concerning surface combustion, depending on the calorific value of the treated waste, the temperature (therefore melting) of the surface layer of the molten material is not always guaranteed, and cooling may occur associated with an accumulation of material that remained solid. Note that a special process (susceptor, metallothermy, etc.) is always necessary when starting direct induction melting processes in a cold crucible when the molten material is not electrically conducting at low temperatures, which is the case for glass.